Most agencies stretch voice AI client onboarding across two to four weeks. With the right sequence, the work inside VoiceAIWrapper, sub-account provisioning and agent sync, takes about 60 minutes for a standard inbound deployment. This playbook walks the five operational steps, what you actually do at each, what goes wrong, and what VoiceAIWrapper adds.
Discovery and scope. Pick the vertical, define one measurable success metric, and confirm the compliance posture before anything gets built.
Agent build inside the provider. Flows, persona, tools, and knowledge base live in Vapi, Retell, ElevenLabs Agents, Bolna, or Ultravox, not in VoiceAIWrapper.
Sub-account provisioning in VoiceAIWrapper (the ~60 minutes). Create the client's branded portal, sync the agent by API key, and set markup pricing.
Telephony and compliance setup. Number provisioning, warm-up cadence, and the consent and disclosure guardrails for the client's jurisdiction.
First live call and monthly reporting cadence. The validation gate, then the monthly report that keeps the client renewing.
If you are evaluating how to onboard a voice AI client, there are real cases where another path serves you better. We will name them. Direct provider setup (Vapi, Retell, or ElevenLabs without any white-label wrapper) is better if you serve a single client and have no plans to resell at scale. Done-for-you agency-implementation consultants (a growing category in the voice AI space) specialize in fully-managed deployments where the agency wants the consultant to handle setup end-to-end rather than running a self-serve software platform internally. Hiring a senior implementation engineer is the right call for enterprise custom builds that need deep CRM integrations, HIPAA BAA at the provider level, or multi-tenant infrastructure that exceeds what a SaaS platform handles. Where VoiceAIWrapper wins: 60-minute sub-account provisioning for standard deployments, sub-account architecture that eliminates per-client infrastructure rebuilds, agency markup pricing built into the billing layer, access to 5 conversational agent platforms in one account (Vapi, Retell, ElevenLabs Agents, Bolna, Ultravox), and monthly reporting tools that give agencies something concrete to invoice against.
WHO THIS PLAYBOOK IS FOR
This page is for agencies that have already decided to offer white-label voice AI and now need the operational sequence to onboard their first or next end client. It covers discovery through monthly reporting. It does not cover market sizing (see Voice AI Market 2026: $47B Agency Capture) or campaign architecture for outbound calling programs (see the outbound AI calling guide). If you are still deciding whether to offer voice AI, start with the market-opportunity page first.
KEY TAKEAWAYS
160-minute setup is a real and specific claim. It refers to sub-account provisioning and agent sync inside VoiceAIWrapper for a standard inbound configuration. Discovery (Step 1) and agent build inside the provider (Step 2) are separate steps that happen before the 60-minute clock starts.
2The 5-step sequence is load-bearing: skip any step and a specific failure mode follows. Discovery without compliance identification leads to mid-project contract renegotiation. Agent build without provider selection logic leads to mismatched provider-to-vertical fit. Sub-account provisioning without markup pricing setup leaves no agency margin.
3Agents configure inside the provider, not inside VoiceAIWrapper. This is the most common architecture misunderstanding new agencies make. Vapi, Retell, ElevenLabs Agents, Bolna, and Ultravox each have their own agent builder. VoiceAIWrapper syncs the completed agent via API key.
4Sub-account architecture is what eliminates per-client infrastructure rebuilds. On Scale tier ($249/month) and above, VoiceAIWrapper supports unlimited client accounts. Each client gets its own branded portal, billing plan, and analytics. The agency's cost-per-additional-client is zero beyond the platform subscription.
5Agency markup pricing is built into VoiceAIWrapper's billing layer. VoiceAIWrapper does not mark up provider voice minutes. The agency sets its own client-facing pricing using VoiceAIWrapper's billing tools. Provider minutes bill directly to the agency's provider account at the provider's rate.
6Telephony warm-up failure is the #1 outbound onboarding failure mode. New outbound numbers must warm up starting at approximately 50 calls per day over 2 weeks. Skipping this results in spam flagging: answer rates drop from 45% to 10% in one week, with 2-week minimum recovery time.
7Monthly reporting is the load-bearing retention lever. SMB B2B SaaS voluntary churn typically runs 4-6% per month, and agency-resold software is most at risk when the agency stops showing monthly results. The agencies that stay on the platform longest are those delivering documented, measurable results to their own clients each month.
8Start with BFSI or dental verticals; add healthcare with HIPAA BAA on Pro tier. Healthcare clients require a signed Business Associate Agreement from VoiceAIWrapper ($499/month Pro tier) plus independent HIPAA configuration at the provider level. Discovering this at Step 3 means renegotiating the contract.
9The 7-day free trial of VoiceAIWrapper gives Scale-tier access ($249/month value), no credit card required. That is enough access to complete Steps 2-5 of this playbook for a real client onboarding.
Ready to run through the playbook with a live client?
The 7-day free trial of VoiceAIWrapper gives Scale-tier access, which supports all 5 providers and unlimited client accounts. No credit card required. Complete Steps 2-5 of this playbook for your first client on the trial. Or see the Voice AI Market 2026: $47B Agency Capture page if you are still building the business case for offering voice AI.
The 5-step voice AI client onboarding playbook for agencies
This playbook reflects how experienced VoiceAIWrapper agencies onboard end clients in 2026. Each step has a clear owner (agency vs. provider platform vs. VoiceAIWrapper), a list of what the agency actually does, and an "Agency operator note" flagging the most common failure point at that step. Steps 2 and 3 use two different systems: the provider builds the agent; VoiceAIWrapper provisions the white-label account. Confusing these two systems is the single most common new-agency mistake.
1
Step 1: Discovery and Scope
Before any technical work begins, the agency runs a structured discovery with the client to identify vertical, use case, success metrics, and compliance constraints. This step takes 1-3 hours, not 2-4 weeks. The 60-minute setup VoiceAIWrapper advertises starts at Step 3, not here.
Identify the client vertical (dental practice, law firm, real estate, HVAC, financial services) to determine which provider best fits the use case.
Determine the use case: inbound (appointment booking, FAQ handling), outbound (lead qualification, follow-up calls), or AI chat (web widget). Each maps to different VoiceAIWrapper plan tiers and provider selections.
Define the conversion outcome: booked appointments per week, cost-per-qualified-lead, or answer-rate percentage. Per Vapi playbook: "Define the primary goal before any technical work. Use three questions: who sponsors it, what triggered it, how will success be measured in 12 months?"
Run the 5 Fit Characteristics assessment: high volume, repetitive and predictable, clear success criteria, strong backend systems, time-sensitive value. Per Vapi playbook: "Miss one and you can compensate. Miss two and the project struggles. Miss three and you're better off not starting."
Identify compliance constraints: HIPAA applies to healthcare clients (requires Pro tier BAA from VoiceAIWrapper at $499/month plus BAA at the provider level); TCPA compliance requirements apply to any outbound campaign.
Agree on success metrics and document them before design begins. This document is what justifies the monthly retainer later.
Agency operator note: Skipping compliance identification at discovery is the most expensive mistake at this step. A dental practice needing HIPAA compliance for healthcare clients requires a Pro tier subscription ($499/month) plus a BAA. Discovering this at Step 3 means renegotiating the contract and losing 1-2 weeks. Identify it here.
2
Step 2: Agent Build Inside the Provider (HARD RULE)
Agents are built inside the underlying conversational agent platform, not inside VoiceAIWrapper. VoiceAIWrapper is not an agent builder. Vapi, Retell, ElevenLabs Agents Platform, Bolna, and Ultravox each have their own agent configuration dashboards, conversation flow builders, and knowledge base tools. The agency uses those to build the agent, then syncs it to VoiceAIWrapper in Step 3.
Vapi: code-first, maximum customization, tool calling, best for complex outbound workflows and technical agency teams.
ElevenLabs Agents Platform: voice + chat + multimodal agents + RAG knowledge base + MCP server support + brand voice customization + multilingual (70+ languages). Pricing from $0.08/min on annual Business plans. See the voice AI market trends and provider benchmarks for the full ElevenLabs positioning.
Bolna: Indic language depth using Sarvam AI integration; right choice for South Asian markets or diaspora verticals.
Ultravox: speech-foundation-model architecture that processes audio directly, relevant for latency-sensitive use cases.
Build conversation flows inside the provider following the Capability Ladder: start at Level 2 or 3 (read-only account access or simple updates), not Level 1 (FAQs only) or Level 4+ (complex multi-step workflows with high failure risk).
Configure tool contracts for any CRM integration. Per Vapi playbook: "The LLM will hallucinate behavior for any gap left undefined in the tool contract." Test the agent inside the provider before provisioning the VoiceAIWrapper sub-account.
Agency operator note: Building on the wrong provider for the use case is the load-bearing error at Step 2. A multilingual dental practice routed to Bolna instead of ElevenLabs Agents Platform misses the multilingual TTS depth. A complex multi-step outbound campaign routed to ElevenLabs instead of Vapi misses the code-first customization. Provider selection at Step 2 locks the technical architecture for the engagement.
3
Step 3: Sub-Account Provisioning in VoiceAIWrapper
This is the 60-minute step. Once the agent is built and tested inside the provider, the agency creates the client's branded portal inside VoiceAIWrapper, syncs the completed agent via API key, and configures agency markup pricing. For a standard inbound configuration, experienced agency operators complete this in approximately 60 minutes.
Create a new client account inside VoiceAIWrapper. On Scale tier ($249/month) and above: unlimited client accounts. On Growth tier ($79/month): up to 15 client accounts. On Starter tier ($29/month): up to 5 client accounts. See VoiceAIWrapper's white-label features for the full tier breakdown.
Sync the completed agent from the provider by entering the client's provider API key into VoiceAIWrapper's integration panel. VoiceAIWrapper reads the agent configuration via the API key.
Set the agency's markup pricing for that client. VoiceAIWrapper does not mark up voice minutes. The agency sets its own per-minute or per-seat pricing using VoiceAIWrapper's billing tools. This is where agency margin is created.
On Growth tier and above: configure Stripe integration to rebill clients directly in-app. On Scale and Pro tiers: configure products and packs (up to 30 on Scale, unlimited on Pro) for recurring client billing. See pricing plans starting at $29/month for the full tier comparison.
Configure the white-label portal: custom domain, agency logo, agency pricing tiers. The client sees the agency's brand, not VoiceAIWrapper.
On Scale and Pro tiers: configure multilingual client portals and advanced client access controls if the use case requires it.
Agency operator note: At Scale tier ($249/month), the agency's cost-per-additional-client is zero beyond the platform subscription fee. All incremental revenue is pure markup on voice minutes and the retainer. That math is the monetization engine. Do not skip pricing configuration at this step.
4
Step 4: Telephony and Compliance Setup
For inbound deployments, telephony setup is straightforward: provision a phone number, configure call routing, and verify consent notices are in place for the relevant state (two-party consent in CA, IL, and 11 other states). For outbound deployments, compliance burden is significantly higher and telephony reputation management is the most common failure mode.
For inbound: provision a phone number through the provider, configure business hours rules and overflow behavior, verify call recording consent notices for applicable state laws.
For outbound: number warm-up cadence. New outbound numbers must warm up starting at approximately 50 calls per day, increasing over 2 weeks. Skipping this results in spam flagging. Answer rates drop from 45% to 10% in one week. Recovery takes 2 weeks minimum. This is the #1 outbound onboarding failure mode. See AI outbound calling campaigns for the full campaign architecture.
STIR/SHAKEN attestation: verify the carrier assigns A-level attestation to outbound numbers. B or C attestation degrades call completion rates.
CNAM registration: register the caller name that displays on recipient phones. Empty or generic CNAM contributes to spam flagging. CNAM processing takes 3-5 business days.
TCPA disclosure scripting: for marketing or sales calls, the agent's opening must state the caller's name and location, confirm prior express written consent or obtain verbal consent, and provide an opt-out mechanism during the call. TCPA compliance requirements for the full regulatory summary.
Do Not Call registry scrub: remove all numbers on the National DNC registry before loading leads. See the AI outbound calling playbook for the full pre-campaign checklist.
Calling window enforcement: configure calling schedule to 8am-9pm local time of the recipient. VoiceAIWrapper's outbound campaign scheduler includes timezone-aware calling controls.
For healthcare clients: Pro tier BAA from VoiceAIWrapper ($499/month) is required before processing any PHI. Provider accounts (Vapi, Retell, ElevenLabs) must be independently contracted for HIPAA compliance. VoiceAIWrapper operates a shared responsibility model.
Agency operator note: Telephony setup for outbound campaigns is where most agencies lose 2-4 weeks. The number warm-up cadence alone takes 2 weeks. Start it on day one of client engagement, not after the agent is built. Parallel-track Steps 2 and 4 for outbound clients.
5
Step 5: First Live Call and Monthly Reporting Cadence
The first live call is the validation gate: confirm the agent answers correctly, CRM round-trips complete, call recording triggers, and call data appears in the analytics dashboard. What happens after the first call determines whether the client stays for 6 months or churns in 60 days. SMB B2B SaaS voluntary churn typically runs 4-6% per month, and agency-resold software is most at risk when the agency stops showing monthly results. The agencies that stay longest are those delivering documented monthly improvements to their own clients.
Run a supervised live test call with the client present. Confirm: agent answers correctly, tool calls to CRM complete successfully, call recording triggers, call data appears in the analytics dashboard.
Verify the CRM round-trip: call outcome writes back to the CRM, lead status updates correctly. See voice AI CRM integration for the round-trip verification checklist.
Establish the monthly reporting cadence in the onboarding contract itself, not as an add-on. Clients expect it from day one if it is in the contract.
Monthly report minimum content: (1) calls completed and answer rate for the month, (2) primary conversion metric vs. the baseline agreed at discovery, (3) one optimization change made this month and its before/after effect, (4) cost per conversion vs. equivalent human labor. The optimization change is what differentiates "running calls" from "managing your voice AI program."
Run the weekly optimization loop: per Vapi playbook documentation, an insurance brokerage moved from 12% to 27% transfer-to-close rate over 14 weeks via one targeted change per week. Each weekly change generates one monthly reporting data point. See improving outbound AI calling performance for the analytics-driven optimization framework.
Agencies that run the monthly reporting loop have client retention data to show. Agencies that stop proactive communication after onboarding see clients cancel 60-90 days later with "what is this for again?"
Agency operator note: The monthly report is the retainer justification. Without it, the agency is billing for uptime, not for results. Clients who see month-over-month improvement in their conversion metric stay. Clients who only receive an invoice without context do not. Build the reporting template at Step 1, populate it from Step 4 analytics, and deliver it on a fixed date each month starting from the first full month live.
Ready to run Steps 2-5 with a real client?
Start the 7-day free trial of VoiceAIWrapper. The trial gives Scale-tier access, which supports all 5 providers, unlimited client accounts, and the full sub-account provisioning flow covered in Step 3. No credit card required. Or review pricing plans starting at $29/month to pick the right tier before signing your first client.
Which of the 5 providers to pick for each client vertical
Provider selection at Step 2 locks the technical architecture for the client engagement. Match the provider to the client's vertical and use case before building the agent. VoiceAIWrapper is not a competitor to any of these platforms; it is the agency monetization and portfolio management layer that syncs agents from all 5 via API key. For provider-specific guides, see the white-label ElevenLabs Agents guide and the Vapi optimization guide for agencies.
Vapi
Best for: complex outbound workflows, technical agency teams, code-first customization. Code-first, open API, STT-LLM-TTS pipeline. Available on Starter, Growth, Scale, and Pro VoiceAIWrapper tiers. VoiceAIWrapper listed as Vapi platform partner.
Retell
Best for: inbound use cases where responsiveness is the primary success metric. Lower median latency than Vapi per Tested Media April 2026 benchmarks. Available on Starter, Growth, Scale, and Pro tiers. Good fit for healthcare inbound (scheduling, triage) and legal intake.
ElevenLabs Agents
Best for: voice quality-sensitive verticals (luxury, medical, executive), multilingual deployments (70+ languages), and multimodal voice + chat + text use cases. Full AI Agents Platform with RAG, MCP, brand voice. Pricing from $0.08/min on annual Business plans. Scale and Pro tiers.
Bolna
Best for: South Asian markets and Indic language deployments (11+ languages via Sarvam AI integration). Indian carrier integrations (Plivo, Exotel, Vobiz) unavailable on other platforms. Scale and Pro tiers. Right choice for diaspora verticals in North America serving South Asian communities.
Ultravox
Best for: latency-sensitive use cases and technical buyers who want speech-foundation-model architecture. Processes audio directly, skipping STT-LLM-TTS intermediate stages for materially lower time-to-first-token. Scale and Pro tiers. Open-weight model at fixie-ai/ultravox.
Start narrow: one vertical, one provider for your first 90 days
Pick one vertical for the first 90 days. Build the agent template, the compliance checklist, and the monthly reporting template for that one vertical. Then expand. Per Vapi's Capability Ladder logic: start narrow, prove the technology works, build operational confidence, then expand. Too many verticals at launch is the second most common onboarding failure mode.
Provider isolation across clients reduces portfolio risk
Different clients can run on different providers inside one VoiceAIWrapper account. VoiceAIWrapper surfaces alerts and analytics when a provider degrades; the agency chooses when to swap. There is no automatic mid-call failover, and that is intentional: the platform does not interfere with agency runtime decisions.
No per-minute markup: provider cost passes through directly
Provider minutes bill directly to the agency's provider account at the provider's rate. VoiceAIWrapper does not mark up those minutes. The agency sets its own client-facing pricing plans using VoiceAIWrapper's billing tools. Confirm current provider rates at each provider's pricing page before scoping a client retainer.
Telephony requirements apply to all 5 providers equally
CNAM registration takes 3-5 business days. STIR/SHAKEN attestation verification takes 1-3 business days. Number warm-up at approximately 50 calls per day over 2 weeks applies regardless of which provider the agent runs on. Start telephony setup at Step 4 in parallel with agent build, not after it.
For comparison:Vapify wraps Vapi only · Voicerr wraps mostly Vapi · ChatDash wraps 3 agent platforms (Vapi, Retell, ElevenLabs) · Synthflow / Insighto / Thinkrr run proprietary engines. VoiceAIWrapper's 5-in-one multi-provider structure is the right choice when an agency needs to match different clients to different providers based on vertical, compliance, or language requirements.
Why the 5-step sequence matters: what the provider handles vs. what VoiceAIWrapper adds
VoiceAIWrapper is NOT an agent builder. Agents configure inside the chosen provider (Vapi, Retell, ElevenLabs Agents, Bolna, or Ultravox); VoiceAIWrapper syncs them via API key and adds the agency monetization and portfolio management layer on top. These two tables show the precise split, so agencies know which dashboard to open at each step.
WHAT HAPPENS INSIDE THE PROVIDER (STEP 2: AGENT BUILD)
Inside the provider (Vapi, Retell, ElevenLabs Agents, Bolna, or Ultravox dashboard)
Knowledge base: RAG document upload, FAQ configuration, structured data retrieval
Inside the provider (varies by platform; ElevenLabs and Vapi have native RAG)
Voice quality and multilingual TTS: voice selection, brand voice customization, language settings
Inside the provider (ElevenLabs: 70+ languages; Bolna: 11+ Indic languages via Sarvam AI)
Tool calling and CRM integration: webhook setup, tool contracts, API connections
Inside the provider (configure tool contracts before syncing to VoiceAIWrapper)
Telephony provisioning: phone number assignment, call routing, batch dialing setup
Inside the provider (Vapi, Retell, ElevenLabs each have native telephony; BYO carrier also supported)
Conversation analytics: call recordings, transcript review, per-agent performance data
Inside the provider (VoiceAIWrapper surfaces portfolio-level analytics; per-agent data lives in provider)
WHAT HAPPENS INSIDE VOICEAIWRAPPER (STEP 3: SUB-ACCOUNT PROVISIONING)
VOICEAIWRAPPER-NATIVE CAPABILITY
Agency markup pricing: set client-facing pricing plans at any markup, currency, and billing frequency; provider cost never visible to the client
VoiceAIWrapper billing tools (Starter and above)
Sub-account architecture: one agency account, many isolated client portals with separate analytics, billing plans, and client logins
VoiceAIWrapper sub-account system (Starter: 5 clients; Growth: 15 clients; Scale and Pro: unlimited)
API key sync: enter the client's provider API key; VoiceAIWrapper reads the completed agent configuration
VoiceAIWrapper integration panel (all tiers)
Pods architecture: attach multiple providers to one client portal; run side-by-side on one account
VoiceAIWrapper Scale and Pro tiers
Phone-number pool: distribute high-volume outbound across multiple numbers
VoiceAIWrapper outbound campaign controls
HIPAA BAA for healthcare-vertical clients: signed Business Associate Agreement from VoiceAIWrapper
VoiceAIWrapper Pro tier ($499/month) only
Architecture rule (HARD RULE): Configure the agent inside the provider (what it says, what it knows, what tools it uses). VoiceAIWrapper syncs it via API key and adds billing, portfolio management, and multi-provider access. It is not an agent builder. See the VoiceAIWrapper feature set for the complete native capability list.
FIRST-HAND · OPERATOR OBSERVATIONS FROM RUNNING AGENCY ONBOARDING
Five things we have seen agencies get wrong in their first 90 days of client onboarding
These observations reflect direct operational experience building and running VoiceAIWrapper across all 5 conversational agent platforms since the platform launched in May 2025, and what the agencies on the platform surface through usage patterns and support conversations.
Why 60-minute setup matters in agency sales conversations
The 60-minute claim is not primarily a technical claim; it is a sales argument. When an agency pitches voice AI to a dental practice, the practice owner is thinking "how long until this disrupts my staff and my systems?" A 60-minute sub-account provisioning step converts a 3-month perceived project into a same-day activation. The agencies on VoiceAIWrapper that close deals fastest are the ones who lead the sales conversation with "your branded portal is live before the end of today's call, and your first call campaign is running by Friday." The 60-minute setup is what makes that statement accurate. Frame it in client terms, not technical terms.
The agent-build-vs-sub-account mistake new agencies make
The most common call to VoiceAIWrapper support in the first 90 days of an agency account: "I built the agent in VoiceAIWrapper but the client portal is not showing the right configuration." The agent was not built in VoiceAIWrapper. Agents build inside the provider. VoiceAIWrapper syncs the completed agent via API key. When agencies try to configure conversation flows in VoiceAIWrapper, they are working in the wrong system. The fix is always the same: go back to the provider dashboard, configure the agent there, then re-enter the API key in VoiceAIWrapper's integration panel. This mistake costs 2-4 hours per new agency operator on average. Reading Step 2 of this playbook before onboarding eliminates it.
Telephony warm-up failure mode: 45% to 10% answer-rate collapse
The Vapi playbook documents a logistics agency deployment where new outbound numbers were loaded at full campaign volume without a warm-up period. Answer rates dropped from 45% to 10% in one week after the numbers were flagged as spam. Recovery took 2 weeks, during which the client's outbound campaign was effectively dead. The agency nearly lost the client. The fix is identical every time: warm up new numbers starting at approximately 50 calls per day, increasing over 2 weeks, before running any full-volume campaign. This is not a VoiceAIWrapper configuration step; it is a telephony hygiene step that applies regardless of provider. But agencies who have not done outbound calling before do not know this.
Monthly reporting cadence is what stops voluntary churn
SMB B2B SaaS voluntary churn typically runs 4-6% per month; agency-resold tools are most at risk when the agency stops showing the client documented monthly results. The pattern is consistent: agencies that stop proactive client communication after the first month see clients cancel 60-90 days post-onboarding. The question is always some variant of "what is this for again?" Agencies that run the monthly reporting cadence described in Step 5, documenting one optimization change per month and its measured effect, have client relationships that compound. The monthly report is the retainer justification. Without it, the agency is billing for uptime, not for results, and clients notice when the next renewal comes.
Vertical-first strategy: start in BFSI or dental, expand to healthcare with HIPAA
Agencies that try to serve every vertical simultaneously in their first 90 days spread thin, deliver mediocre results across all accounts, and see clients churn across all of them. The agencies on VoiceAIWrapper that grow fastest pick one vertical for the first 90 days, build the agent template and compliance checklist for that vertical, and only then expand. BFSI and dental are the lowest-compliance-burden entry points. Healthcare requires a Pro tier BAA from VoiceAIWrapper ($499/month) plus independent HIPAA configuration at the provider level. Discovering this at Step 3 after a healthcare client is already signed is the onboarding equivalent of building the wrong provider selection at Step 2. Identify compliance constraints at Step 1.
These observations reflect operational patterns from running VoiceAIWrapper since May 2025, not individual case studies from specific named customers. For the agency revenue and pricing framing, see the companion page: Voice AI Market 2026: $47B Agency Capture . Authored by Raj Baruah, Founder, VoiceAIWrapper. 8+ years in B2B SaaS. Adobe alum.
COST LAYERS DURING ONBOARDING
Cost layers an agency carries during client onboarding
An agency's cost structure during voice AI client onboarding has four distinct layers. Understanding each layer before Step 3 prevents margin miscalculation and cost-plus pricing collapse (Failure Mode 1 in Section 8). VoiceAIWrapper does not mark up provider minutes; those costs pass through directly to the agency at the provider's rate. All rates below must be verified against each provider's live pricing page before use in any client proposal; rates change independently of this page.
Provider voice minutes (example: Vapi platform fee) Verify at Vapi Pricing Page; rates change
~$25-75/mo at provider rate
~$125-375/mo at provider rate
~$500-1,500/mo at provider rate
~$1,250-3,750/mo at provider rate
Telephony (phone number rental + carrier fees) Varies by provider and country
~$5-15/mo per number
~$25-75/mo
~$100-300/mo
~$250-750/mo
Agency billed price to client Set by agency; this is where margin lives
Agency sets (e.g., $150-500/mo)
Agency sets (e.g., $500-2,500/mo total across clients)
Agency sets (e.g., $2,000-10,000/mo total)
Agency sets (e.g., $5,000-25,000/mo total)
Methodology: Provider per-minute rates are not hard-coded here because they change as platforms scale. The VoiceAIWrapper subscription costs are from the pricing snapshot. Provider minute costs are illustrative ranges and must be verified at each provider's live pricing page before any client proposal. VoiceAIWrapper adds no per-minute markup; provider costs pass through directly. The agency's billed price to the client is set by the agency using VoiceAIWrapper's billing tools. For the full agency markup math and revenue scenario model, see Voice AI Market 2026: $47B Agency Capture.
VOICEAIWRAPPER TIER COMPARISON FOR ONBOARDING
Which VoiceAIWrapper tier covers which onboarding capability
The right VoiceAIWrapper tier depends on your client count, the number of providers you need, and whether your clients are in healthcare (HIPAA BAA required). Choose before Step 3. See the Vapi optimization guide for agencies and the white-label ElevenLabs Agents guide for provider-specific configuration depth.
Starter: $29/month · Annual $24/mo
Client accounts: Up to 5. Providers: Vapi + Retell (2 of 5). Campaigns: Up to 5. Stripe billing: Not included; invoice clients manually. Right for: New agencies onboarding their first 1-3 clients. Complete Steps 2-5 of this playbook. Test the platform before upgrading.
Growth: $79/month · Annual $66/mo
Client accounts: Up to 15. Providers: Vapi + Retell (2 of 5). Campaigns: Unlimited. Stripe billing: Included; rebill clients directly in-app. Smart Callback: Included. Right for: Agencies actively billing 5-15 clients through the platform. Growth is the right tier once the agency is charging clients through VoiceAIWrapper rather than invoicing manually.
Scale: $249/month · Annual $208/mo · Most Popular
Client accounts: Unlimited. Providers: All 5 (Vapi + Retell + ElevenLabs Agents + Bolna + Ultravox). Campaigns: Unlimited. Products and packs: Up to 30. Multilingual portals: Included. Phone-number pool: Included. Right for: Agencies running multi-provider stacks or international/multilingual campaigns. The 7-day free trial gives Scale-tier access.
Pro: $499/month · Annual $416/mo
Client accounts: Unlimited. Providers: All 5. Products and packs: Unlimited. HIPAA BAA: Included (required for healthcare clients with PHI). Custom development: Available. Right for: Enterprise agencies serving healthcare, financial services, or other regulated verticals that require a signed Business Associate Agreement.
✅5 voice providers under one dashboard
VoiceAIWrapper Provider Count
1VoiceAIWrapper - 5
2ChatDash - 3 (Vapi, Retell, ElevenLabs)
3Synthflow Agency - 1 (proprietary)
4Vapify - 1 (Vapi only)
5Voicerr - Vapi-focused
6Direct provider - 1 provider per account
COMMON FAILURE MODES
Three onboarding failure modes that drive client churn (and how to avoid them)
SMB B2B SaaS voluntary churn typically runs 4-6% per month; for agency-resold software the bigger risk is the agency's own client losing interest. The failure modes below are the proximate causes. Agencies that fix them retain clients. Agencies that do not lose them within 60-90 days of onboarding.
FAILURE MODE
WHAT HAPPENS
WHY IT HAPPENS
THE FIX
Cost-plus pricing collapse Most common among new agency operators
Agency prices at cost-plus. When the client's call volume grows, the agency's margin compresses. The client compares prices and cancels.
Pricing on cost (VoiceAIWrapper subscription + provider minutes) rather than the value delivered to the client.
Price on outcome. An appointment-setting agent generating 30 additional dental appointments per month at $80 average revenue creates $2,400 in client revenue. A $300-500/month retainer against that outcome is defensible.
Too many verticals at launch Spreads operational knowledge too thin
Agency offers voice AI to dental, law, real estate, HVAC, and e-commerce simultaneously. Quality suffers across all accounts. Clients churn from multiple verticals at once.
Voice AI is horizontal technology. The value proposition is vertical-specific. Each vertical requires different compliance knowledge, conversation flows, and success metrics.
Pick one vertical for the first 90 days. Build the agent template, compliance checklist, and monthly report template for that one vertical. Expand after you have a proven repeatable process.
No monthly reporting cadence The root cause of most 60-90 day churn
Onboarding completes, the agent goes live, and the agency stops proactive communication except when something breaks. Client cancels 60-90 days later asking "what is this for again?"
Agencies treat voice AI as a one-time setup. Clients treat it as an ongoing investment that needs to show ROI. Without regular reporting, clients fill the uncertainty gap with doubt.
The monthly report is the retainer justification. Without it, the agency bills for uptime, not for results. The Vapi weekly optimization loop (one targeted change per week, documented) gives the agency a concrete data point to report every month.
Data note: The 4-6%/month figure is the published SMB B2B SaaS voluntary-churn benchmark. The failure modes above represent the proximate causes of agency-level attrition: agencies that cannot retain their own clients drop the platform. For the onboarding mechanics that prevent each failure mode, see the outbound AI calling performance guide and the full AI outbound calling use case page.
QUARTERLY IMPROVEMENT CADENCE
Quarterly onboarding improvement cadence: what to review every 90 days
A well-onboarded client on day 1 is not the same as a retained client at month 6. Agencies that grow their voice AI book of business run structured quarterly improvement reviews across four dimensions. This cadence is the operational version of the weekly optimization loop, zoomed out to the quarterly level.
Q1 Review (Month 3)
Discovery quality audit
Were the success metrics agreed in Step 1 actually measurable in the first quarter's data? If not, renegotiate the metrics before Month 4 billing.
Did the 5 Fit Characteristics assessment (high volume, repetitive, clear success criteria, strong backend systems, time-sensitive value) hold up in practice? Document any gaps.
Has the client's call volume hit the telephony warm-up ceiling? Review whether number expansion is needed.
Is the agent still operating at the right Capability Ladder level? Most agents should be at Level 2-3 after 90 days. Consider moving up only if the Level 2-3 outcomes are stable.
Q2 Review (Month 6)
Telephony reputation check
Pull answer rate data for outbound clients. Has answer rate held above 35-45%? If it has dropped below 20%, run the spam-flag diagnosis: CNAM registration, STIR/SHAKEN attestation level, call cadence.
For inbound clients: review abandonment rate. If more than 15% of calls end before the agent's second turn, review the agent's opening prompt.
Review any carrier-level complaints or flagged numbers. Address before Quarter 3 volume increases.
Confirm STIR/SHAKEN attestation is still at A-level for all outbound numbers. B or C attestation degrades completion rates.
Q3 Review (Month 9)
Pricing structure audit
Review the client's current retainer against their call volume and outcome data. Is the agency still pricing on cost-plus, or have they moved to outcome pricing? Month 9 is the natural renewal window.
Calculate cost per conversion for the last quarter. If cost per conversion has fallen, the agency has a concrete case for renewing the retainer without reduction. If it has risen, identify the cause before the renewal conversation.
Review VoiceAIWrapper plan tier: is the agency on the right tier for their current client count? Growth ($79/mo) caps at 15 clients; Scale ($249/mo) is unlimited.
Q4 Review (Month 12)
Vertical expansion or contraction decision
Run a formal audit of which vertical is delivering the best client retention and the best agency margin. The first vertical the agency onboarded should by now have a repeatable template: agent build, compliance checklist, monthly report.
Decide: expand to a second vertical using the same template, contract to focus only on the best-performing vertical, or deepen the current vertical with a higher Capability Ladder tier.
Review whether multi-provider configurations (ElevenLabs + Vapi, or Bolna for Indic markets) add margin for specific client types. Scale and Pro tiers include all 5 providers.
For agencies scaling past a handful of clients: Year 1 completion is the natural point to move onto the Scale and Pro plans , which unlock unlimited client accounts and all five providers.
Why quarterly cadence matters: In SMB B2B SaaS, voluntary churn typically runs 4-6% per month, and the agencies most exposed are those that do not run structured improvement reviews. Clients who see quarterly evidence of improvement in their conversion metrics stay. Clients who only receive invoices do not. The quarterly cadence is the agency's answer to the client's question: "Is this still worth the spend?"
AGENCY ONBOARDING CHECKLIST
Agency onboarding checklist: 10 steps before the first live call
Use this checklist before the first live production call with any new client. Each item maps to a section of this playbook. Running through the list takes approximately 30 minutes. It prevents the most common onboarding failures.
1
Discovery: 5 Fit Characteristics assessed and documented
Complete the Vapi Capability Ladder fit assessment for the client's use case: high volume, repetitive and predictable, clear success criteria, strong backend systems, time-sensitive value. Per Vapi playbook: "Miss one and you can compensate. Miss two and the project struggles. Miss three and you're better off not starting." Record which of the 5 the client satisfies and which they do not. This document lives in the client file and justifies the retainer if the engagement is challenged later.
2
Discovery: success metric defined in the client's own terms
Before any build work begins, document the single primary metric the client will use to judge success at Month 3. Examples: appointments booked per week, qualified leads per day, answer rate percentage for inbound calls. This metric drives the monthly report. Without a defined baseline, the report has no reference point.
3
Discovery: compliance requirements identified early
Healthcare clients require a HIPAA BAA (Pro tier, $499/mo from VoiceAIWrapper). Outbound campaigns require TCPA compliance planning (prior express written consent, DNC registry scrub, 8am-9pm calling window). Identify these before Step 2 agent build. Discovering a HIPAA requirement at sub-account provisioning means renegotiating the contract.
4
Provider selected: correct platform for the use case
The agent is being built in the right provider for the client's use case: Vapi for complex outbound workflows, Retell for inbound responsiveness, ElevenLabs Agents for multilingual or multimodal needs, Bolna for Indic language markets, Ultravox for latency-sensitive deployments. Provider selection at this step locks the technical architecture for the engagement.
5
Agent build: Capability Ladder Level 2-3 validated inside the provider
The agent handles Level 2 or 3 tasks: read-only account lookups or simple update operations, not FAQ-only (Level 1) and not complex multi-step workflows (Level 4+). The agent has been tested inside the provider's own interface and passes basic functional validation before sub-account provisioning begins in VoiceAIWrapper.
6
VoiceAIWrapper sub-account: client portal created and agent synced
The client sub-account is provisioned in VoiceAIWrapper. The provider API key is entered and VoiceAIWrapper has read the agent configuration. The client's portal shows the agency's logo and brand, not VoiceAIWrapper. Markup pricing is set. On Growth tier and above: Stripe rebilling is configured. Estimated time for this step on a pre-built Vapi or Retell agent: approximately 60 minutes.
7
Telephony: phone number provisioned and warm-up started (outbound)
For outbound campaigns: new numbers are starting at approximately 50 calls/day and will scale over 2 weeks. CNAM registration is confirmed. STIR/SHAKEN attestation is verified at A-level. Do Not Call registry scrub is complete for the first lead batch. Calling hours are configured for 8am-9pm recipient local time.
8
Compliance: consent documentation in place for outbound; call recording notices for inbound
For outbound: proof of prior express written consent exists for the lead list being called. The agent's opening statement includes caller identification and an opt-out mechanism. For inbound: call recording consent notice is configured for any two-party-consent state (California, Illinois, and 11 others).
9
First live call: supervised test with client present
Run a supervised test call with the client watching (or on the call). Confirm: agent answers correctly per the conversation flow, tool calls to the CRM complete and return expected data, call recording triggers, call data appears in the VoiceAIWrapper analytics dashboard. Confirm the CRM round-trip: call outcome writes back to the CRM, lead status updates. Document the test outcome in the client file.
10
Monthly reporting: template agreed in the onboarding contract
The monthly report format is agreed with the client before or during onboarding, not after the first invoice. The template covers: calls completed and answer rate, primary conversion metric vs. baseline, one optimization change made this month and its measured effect, cost per conversion vs. equivalent human labor. Agencies that establish this in the contract have a monthly deliverable that justifies the retainer. Agencies that do not are billing for uptime.
WHEN THIS PLAYBOOK DOES NOT FIT
Honest concession: four situations where this playbook is not the right fit
Four situations where this playbook is not the right fit
1You are a solo developer building one custom integration for one enterprise client. This playbook assumes an agency model: multiple clients, repeatable processes, sub-account architecture, and markup pricing. If you are building a single bespoke voice AI deployment for one enterprise client, the sub-account provisioning and markup pricing mechanics in Steps 3-4 are not relevant. Custom enterprise builds are better served by direct Vapi or ElevenLabs Agents Platform implementation guides. For Vapi configuration at the implementation level, see the Vapi optimization guide for agencies.
2Your client's only use case is pure inbound with no outbound component. Steps 4's telephony warm-up doctrine and TCPA compliance section are written for outbound campaigns. If your client exclusively runs inbound voice AI (a dental practice using an AI receptionist, for example), the number warm-up cadence, STIR/SHAKEN attestation checks, and CNAM registration steps do not apply. Still run the compliance section's inbound items: call recording consent notices in two-party-consent states and HIPAA review for healthcare clients.
3You need a fully custom-built platform, not a white-label layer. VoiceAIWrapper provides white-label infrastructure for agencies. It does not replace a fully custom-built proprietary voice AI platform. Agencies that have already invested in custom infrastructure (their own telephony stack, their own sub-account database, their own billing engine) will not see the same time-to-market benefit from the 60-minute setup. This playbook's value is highest for agencies that have not built that infrastructure themselves and do not want to.
4You are a single-vertical specialist agency already at $5M+ revenue in that vertical. At significant scale within one vertical, you have likely already built proprietary compliance templates, conversation flow libraries, and reporting systems. The playbook's Section 3 (sub-account provisioning) and Section 4 (telephony setup) may replicate work you have already done in your own infrastructure. At that scale, the relevant question is whether a multi-provider white-label layer (5 conversational agent platforms in one account) reduces your cost per client acquisition or adds capability for adjacent verticals you want to enter, not whether the basic onboarding mechanics apply.
Ready to run your first client onboarding in 60 minutes?
The 7-day free trial gives you Scale-tier access (unlimited clients, all 5 providers, multilingual portals). No credit card required. Start the onboarding checklist from Step 3 on day one.
Four objections to the onboarding playbook (and honest answers)
""60 minutes is not realistic for a healthcare client with HIPAA requirements."Correct, and the playbook says so explicitly in Step 1. The 60-minute setup refers to sub-account provisioning and agent sync inside VoiceAIWrapper (Step 3) for a standard configuration. Healthcare clients require additional time: a HIPAA BAA from VoiceAIWrapper (Pro tier, $499/mo), plus independent HIPAA configuration at the provider level (Vapi, Retell, or ElevenLabs each require their own BAA). The full onboarding for a healthcare client runs 1-2 weeks minimum for an inbound setup, and 3-4 weeks for outbound with full consent documentation. The 60-minute signal is for non-regulated inbound deployments. It is not a universal promise for all verticals.
""I already have my Vapi account set up. Why do I need VoiceAIWrapper on top?"Having a Vapi account means you have an agent-building environment. It does not give you the agency-layer infrastructure this playbook is about. VoiceAIWrapper adds: a separate white-label portal for each client (the client sees your brand, not Vapi), agency markup pricing and Stripe rebilling configured in-app, a sub-account structure that isolates each client's data and analytics, and access to 4 additional providers (Retell, ElevenLabs Agents, Bolna, Ultravox) in the same account. If you are billing one client for one Vapi deployment, you may not need VoiceAIWrapper yet. If you are managing 3+ clients and want to charge markup on voice minutes without rebuilding billing infrastructure, that is what Step 3 provides.
""Monthly reporting feels like overhead that slows down the agency."Monthly reporting takes approximately 30-60 minutes per client using call analytics data already captured in VoiceAIWrapper and the provider dashboard. The question is not whether reporting takes time (it does). The question is whether the 30-60 minutes of reporting is the thing that keeps the client billing for another 90 days. Agencies that do not run this cadence are the ones that lose clients 60-90 days post-onboarding. Clients who see month-over-month improvement data stay. Clients who only receive invoices cancel. The overhead calculates as: 1 hour per client per month versus the revenue impact of losing the client.
""Do I really need agency markup pricing in software, or can I just manage it on my own invoices?"You can manage client billing outside VoiceAIWrapper with separate invoicing software. The reason to use in-app markup pricing (available on Growth tier and above) is not accounting convenience. It is the client experience. When clients log into their white-label portal and see their usage, minutes consumed, and invoice in the same interface, it reduces billing questions and disputes. The Growth tier Stripe integration ($79/mo) lets clients self-serve their usage data. That self-service capability is part of what justifies the retainer as a productized service rather than a freelance engagement.
Frequently Asked Questions
Question
How long does voice AI client onboarding take?
Answer
For a standard agency setup using VoiceAIWrapper, sub-account provisioning and agent sync takes approximately 60 minutes. The full onboarding process (discovery, agent configuration inside the provider, telephony setup, and client training) typically runs 1-2 weeks for a straightforward inbound use case. Outbound campaigns with compliance requirements (TCPA, HIPAA) add 1-2 weeks for number warm-up and consent documentation. Compare this to traditional custom voice AI development, which takes 3-6 months and $15,000+ upfront.
Question
What does the VoiceAIWrapper 60-minute setup actually include?
Answer
The 60-minute setup refers specifically to sub-account provisioning and agent sync inside VoiceAIWrapper: the actions in Step 3 of the onboarding process. The agency creates the client's branded portal, enters the provider API key, VoiceAIWrapper reads the agent configuration, and the agency sets markup pricing. For a pre-built agent on Vapi or Retell, this step can be completed in approximately one hour. Discovery (Step 1) and agent configuration inside the provider (Step 2) are separate and happen before the 60-minute clock starts.
Question
How do I price voice AI to a client without commoditizing?
Answer
Price on the outcome the client values, not on your cost. An appointment-setting agent that generates 30 additional bookings per month at $80 average value creates $2,400 in client revenue. A $300-500/month retainer is defensible against that outcome. Cost-plus pricing (marking up your VoiceAIWrapper subscription cost and passing through voice minutes at cost) leaves agencies with no margin as client call volume grows and no story to tell when clients compare prices. The monthly report (Step 5) is the proof document that justifies the outcome price.
Question
How do I pick which vertical to onboard first?
Answer
Pick the vertical where you already have a client relationship, understand the compliance requirements, and can define a measurable success metric before the first call. Healthcare and legal clients require HIPAA or specific compliance postures. Start with a less compliance-heavy vertical if you are new to voice AI onboarding. Within the vertical, run the 5 Fit Characteristics from Vapi's voice agent playbook: high volume, repetitive and predictable, clear success criteria, strong backend systems, time-sensitive value. Miss three of five and the use case is not yet ready for voice AI.
Question
What is the difference between agent build and sub-account setup?
Answer
These are two separate steps in two different systems. Agent build happens inside the underlying provider (Vapi, Retell, ElevenLabs Agents, Bolna, or Ultravox): this is where conversation flows, personas, tool connections, and knowledge bases are configured. Sub-account setup happens inside VoiceAIWrapper: this is where the agency creates the client's branded portal, syncs the completed agent via API key, and sets agency markup pricing. VoiceAIWrapper does not build agents; it makes completed agents resellable under the agency's brand.
Question
How do I handle TCPA compliance for an outbound client?
Answer
TCPA compliance for outbound voice AI requires: prior express written consent before calling mobile numbers with an automated dialing system, a Do Not Call registry scrub before any campaign, calling only between 8am and 9pm in the recipient's local timezone, clear caller identification in the agent's opening statement, and an opt-out mechanism during the call. New phone numbers must be warmed up gradually (start at approximately 50 calls/day, increase over two weeks) before running full campaigns. Skipping warm-up results in numbers being flagged as spam, dropping answer rates from 45% to 10% in one week. See the TCPA compliance requirements page for the full regulatory summary.
Question
What should monthly client reporting include?
Answer
A minimum monthly report includes: (1) total calls completed and answer rate for the month, (2) primary conversion metric versus the baseline agreed at discovery (appointments booked, leads qualified, cost per conversion), (3) one optimization change made this month and its measured before/after effect, and (4) cost per conversion versus equivalent human labor cost. The optimization change is what differentiates a "we are just running calls" agency from a "we manage and improve your voice AI program" agency. The latter can defend a retainer; the former cannot.
Question
Which VoiceAIWrapper plan do I need to onboard my first client?
Answer
The Starter plan ($29/mo) supports up to 5 client accounts, Vapi and Retell integrations, and up to 5 campaigns. For a single first client with a basic inbound or outbound use case, Starter is sufficient. Growth ($79/mo) adds Stripe integration for in-app client billing, unlimited campaigns, Smart Callback, and up to 15 clients. That is the right tier once the agency is billing clients through the platform. Scale ($249/mo) adds ElevenLabs Agents, Bolna, and Ultravox, plus unlimited clients and multilingual portals. That is the tier for agencies running multi-provider stacks or international/multilingual campaigns. See all plan details at VoiceAIWrapper pricing.
Raj Baruah, Founder, VoiceAIWrapper
Raj built VoiceAIWrapper to give agencies the sub-account architecture, agency markup billing, and multi-provider white-label layer they would otherwise have to build from scratch on top of Vapi, Retell, ElevenLabs Agents, Bolna, and Ultravox. Because VoiceAIWrapper aggregates all 5 conversational agent platforms in a single operator account, Raj observes the market from a position that no single-provider analyst or operator has: what different provider architectures reveal about market direction, which latency and compliance thresholds trigger client decisions, and how per-minute cost structures interact with agency margin across different verticals. The market trends on this page reflect that multi-platform operational perspective, layered on top of the named primary research sources.
For the agency monetization angle (how to price a retainer, which provider to pick per vertical, what VoiceAIWrapper's sub-account architecture costs at different agency sizes), see Voice AI Market 2026: $47B Agency Capture. Healthcare-vertical agencies should review the HIPAA compliance posture before scoping client retainers.
LinkedIn: rajbaruahListed Vapi platform partnerVoiceAIWrapper LinkedInFeatured expert: Raj Baruah on ConnectivelyVoiceAIWrapper Academy community on Skool5.0/5 on SaaSHub (17 verified reviews)
Found our insights helpful? Start your voice AI white label free trial
Our product is free to use for 7 days (no credit card required). You get access to premium features available in our Scale plan during your free trial.
If you are not satisfied with our product or support, we offer you a full refund. For details, please read our refund policy in the footer of our home page.
Used by 1000+ agencies.
99.9% uptime.
60-minute setup.
Found our insights helpful? Start your voice AI white label free trial
Our product is free to use for 7 days (no credit card required). You get access to premium features available in our Scale plan during your free trial.
If you are not satisfied with our product or support, we offer you a full refund. For details, please read our refund policy in the footer of our home page.
Used by 1000+ agencies.
99.9% uptime.
60-minute setup.
Found our insights helpful? Start your voice AI white label free trial
Our product is free to use for 7 days (no credit card required). You get access to premium features available in our Scale plan during your free trial.
If you are not satisfied with our product or support, we offer you a full refund. For details, please read our refund policy in the footer of our home page.
